Complications of mild traumatic brain injury in Veterans and military personnel: A Systematic Review
March 26, 2013
This is an unedited transcript of this session. As such, it may contain omissions or errors due to sound quality or misinterpretation. For clarification or verification of any points in the transcript, please refer to the audio version posted at www.hsrd.research.va.gov/cyberseminars/catalog-archive.cfm or contact firstname.lastname@example.org. Moderator: I am going to do a quick introduction of our speakers and discussants for today’s session. Maya O’Neil is a psychologist and investigator at the Portland VA Medical Center and Assistant Professor at Oregon Health and Science University. She works with Evidenced-based Synthesis Program, HSR&D Research Enhancement Award Program, AHRQ Scientific Resource Center and Neuropsychology Service at the Portland VA. Her research focuses on consequences and treatment of polytrauma as well as systematic review methods.
Kathleen Carlson is an epidemiologist and health services researcher at the Portland VA Medical Center. Her work focuses on unemployment and other functional outcomes among Veterans with TBI and comorbid mental health disorders.
David Cifu is the Chief of PM&R Services at the VCU Health System, the Executive Director of the VCU Center for Rehabilitation Sciences in Engineering and National Director of PM&R Program Office and a member of the Senior Executive Staff to the Department of Veteran’s Affairs.
Daniel Storzbach is a staff psychologist, Program Manager of Mood Disorders Research and Treatment Center and an Associated Professor of Psychology Department of Psychiatry at OHSU.
Joel Scholten is the Washington, DC lead for the Joint Incentive Fund Project for Amputee Care in the National Capital region. He also works in the VA Central Office within the Physical Medicine and Rehabilitation Program Office as the National Director of Special Projects.
Robert Ruff is the Neurology Service Chief at the Louis V. Stokes Cleveland Veteran’s Affair Medical Center and the National Director for Neurology Department of Veteran’s Affairs.
With that Maya I am going to turn things over to you.
Maya O’Neil: Okay, great. I will take myself off mute and we will get started. First of all thanks everyone for joining us this morning. It is TBI Awareness Month so an appropriate topic and thanks for coming. I am going to scroll through some of these introductory slides pretty quickly. Here we have a lot of just general introductory information. In case you do not know these slides are available online for people who cannot stay for the entire presentation and/or to view the presentation later, so we like to put a lot of information in there so you can read through them later at your convenience.
First of all acknowledgements, and this is something that I take pretty seriously. I think it is pretty important. For those of you who are not familiar with systematic reviews they are really, really intensive. They take a lot of work, a lot of effort from a lot of different people. Not only are the report authors very important, particularly our stakeholders. We have technical expert panel and then we have quite a few peer reviewers. So I want to make sure that we…hold on a second. We are getting a couple of questions that people cannot hear. So I am going to put the phone closer. Heidi can you hear me okay?
Moderator: I can you hear you okay but if anyone in the audience is still having issues just send that in so we can see. Thank you.
Maya O’Neil: Just so people know the process of this, I can see the questions that come in. Heidi is going to be doing a lot of the responding as I am talking but we still have some people who cannot hear. Is this significantly better for the folks that cannot hear because this way I am actually picking up the phone instead of talking to the speakerphone? Heidi does that sound better to you?
Moderator: It sounds a little bit clearer. If anyone in the audience…I am hearing no that is not better. If you are on a computer and you are using your computer speakers you may need to dial in on the telephone. You will have a significantly better audio quality. The computer, the speakers that come on like a CPU unit are usually not great quality and you may need to dial in.
Maya O’Neil: Okay. I am going to try putting again on speaker and we will try to move forward with that but if people still have problems hearing please do write in a question and let us know.
Alright, hopefully people can still follow along with the slides even if they cannot hear everything perfectly at this point. Heidi I will let you handle the responses to questions of folks who are still having trouble if that is okay.
Moderator: Sounds good, thanks.
Maya O’Neil: Okay, so back to the acknowledgement side. I want to emphasize not only the focus that we have up here on the slide in front of us but like I said we had quite a few stakeholders for this report. A few of them are on the call today as our discussants. So Doctors Cifu, Hoffman, O’Maya, Ruff and Scholten were all stakeholders for this report. They were the people who initially request the report so I am going to have them talk a little bit about the process of why they requested the report, what information they were hoping to get from it at the end when we have more of a discussion from our stakeholders. It is great that they could be on the call with us today.
We also had quite a few people on our technical expert panel and people who did pretty extensive peer review of our very lengthy report midway through. So those people in addition to the stakeholder panel were Doctors Belanger, Carrol, Eapen, Fann, Frank, Harris, Pogoda, Thayer, Vanderploeg and Vasterling. I just have to say they all provided really extensive comments and guidance as we were writing this report. It is a pretty complicated topic. It is a pretty hot topic in the VA and there are aspects of it that are controversial. So it was very helpful to have extensive guidance as we went through.
Okay, just some disclosure information that we have to put up there that you can read through. We also wanted to tell you for those of you particularly who are not familiar with the Evidence Synthesis Program we wanted to give you a little bit of information about the ESP and what we do. We are sponsored by QUERI. Basically what we do is provide evidence reviews or systematic reviews, Evidence Synthesis Reports to address healthcare topics that are really important to the VA in particular. So we are all affiliated with Evidence-based Practice Centers. There are four of us Evidence Synthesis Programs nationwide. So Evidenced-based Practice Centers are the United States’ way to address some of these important healthcare topics but then our role in the four ESPs are to address questions that are particularly important to Veterans and members of the military.
There is a link there for the nomination process. We take nominations from anyone and everyone, clinicians, folks in central office, etcetera. We have a great coordinating center who are very responsive. If you are interested, if there are clinical questions of interest and you would like people to assist with investigating a topic they will kind of work up a topic for you, see if it warrants a full systematic review and help coordinate that process. So there is the link there if you are interested.
We have an extensive steering committee like we talked about. We have a technical advisory panel; both for the overall ESP and for our individual reports. As I mentioned a lot of peer reviewers and policy partners and we will talk a bit more about that in the discussion portion.
Here are a couple of examples of our recent reports. These are the reports that the Portland VA did in the past fiscal year. Most importantly we have a link. For those of you who are in the VA system you will be able to access the reports immediately. Some of the reports are released intranet only for about six months while they are in the publication process. That is the case with this particular report. So if you are not on a VA computer you will not be able to access the full report for this TBI report for another few months probably. But all of the reports are available there on that link.
Just a bit of an overview of today’s presentation. We are going to keep the background really brief, talk a bit about the scope and methods. We are going to emphasize the results of the report, what we found for these different clinical topic areas and then we are going to have a pretty extensive discussion at the end. Like I said, we have our central office stakeholders on the call with us today. Especially because of some of the limited evidence that was found in our report, it is important to discuss ways to move forward and provide the best care to Veterans and members of the military regardless of some lower strength of evidence. So they are going to help us talk about that.
So a little bit of background. Just very briefly, I am sure those of you who are familiar with the research that is out there on mild TBI, particularly as it relates to OEF, OIF, and OND members of the military and Veterans. A lot of different percentages are thrown about. Generally people talk about 10-23% of service members experiencing a TBI while deployed but those numbers do vary. There are also really differing accounts of mild TBI recovery. That is a lot of what we are going to be talking about in the report today. So existing research really differs in some of those estimates of the different postconcussive symptoms that people experience and how long they last.
A couple of things to point out, one of the reasons that this specific report was requested for Veterans and members of the military is because recovery is likely pretty unique for our OEF, OIF and OND service members. This is because many of them experience multiple mild traumatic brain injuries. The mechanism of injury differs from a lot of the folks in civilian populations who experience a TBI, so this is not your standard athletic event obviously; people who experienced blast exposures, sometimes multiple. And then other physical and mental health concerns. It is likely that PTSD is more common in military settings in people that experience traumatic brain injuries than in a lot of civilian settings. So these are all complicating factors and that is why this particular report was requested with a very specific focus on Veterans and members of the U.S. military.
So here are key questions. Like I said specific to members and Veterans of the military. We were looking at the prevalence of health problems, cognitive deficits, functional limitations, mental health symptoms that develop or persist following an mTBI. Our key question two was really focused on anything that might moderate or mediate any of those mTBI outcomes. So what that means is that we looked at any article that reported on maybe demographic characteristics, comorbid mental health concerns, anything that might possibly affect those outcomes. We tried to take a look at those articles as well. We will talk about that a bit when we get to the results.
Then we were also particularly interested in the resource utilization over time for Veterans and members of the military who have mild TBI. This was primarily for planning purposes. Again we will have our stakeholders talk about this at the latter portion of the talk today.
So a couple of things about methods. I am not going to go into a whole bunch about systematic review methodology but I do want to talk for a bit about how we did this report. A systematic review for those of you who do not know it is very different from a traditional literature review. This is not like people sit down and come up with articles of interest to them or things that they might be familiar with from their colleagues that they know and things like that; that people normally do for a traditional literature review. But rather a very large scale systematic search. So in this case we searched multiple databases including Medline, PsychINFO and the Cochrane Register of Controlled Trials. All of that was searched until relatively recently. Our final search state as we were writing the report was October 3, 2012. So we know that there are articles published since then but for this report that is how far we searched and those are the articles that are included.
Overall it means we searched over 2,600 titles and abstracts. Then 353 of those, after we went through those abstracts and titles, we pulled 353 full text articles to see which one of those met our specific inclusion and exclusion criteria, which I will get to in a second. Thirty-one of those met our inclusion criteria and were included in the report. So that is what we are mostly going to be talking about today, though we will highlight some of the groups of studies that we excluded for certain reasons.
Once those studies were included, we did really extensive quality assessments of all of the primary studies and systematic reviews. We ended up not including any other systematic reviews in this report. So Kathleen Carlson, one of my co-presenters and co-authors on this report was the lead investigator on a recent systematic review talking about PTSD and TBI. So she is going to highlight some of those findings. It was a bit of a different research question but still relevant to the questions we were looking at here.
When we talk about quality assessment, it is pretty important to understand what that means. Quality assessment particularly of observational studies is relatively complicated and really specific to clinical topics. In this case we did not have any randomized controlled trials that met our inclusion criteria and this is not surprising. As you can imagine it would be difficult if not impossible to do ethical and diffusable randomized control trials on TBI. So we had a lot of observational studies.
So we looked at a whole bunch of different quality criteria when we were examining the different observational studies and we will talk about that as we talk about the different results sections, because those quality criteria are very specific to the different studies. So when we talk about studies that are methodologically flawed, one thing I want to emphasize is that this is not an attempt to sort of bash the researchers or say they did not do a good job. In fact, all three of us sitting in the room with me today do TBI research and we know how hard it is to conduct really good quality studies. What I want to emphasize is that a lot of times you are really limited by the populations at hand, the different clinical topics, etcetera. So even if we cannot design a perfect study we still have to be able to look at the information gleaned from these less than perfect observational studies that are still well conducted by the researchers that are doing these studies. Sometimes we just cannot do those ideal studies that we want to see.
Inclusion criteria: Like I said we were very specifically focused on Veterans and members of the military who have experienced a mild traumatic brain injury. One of the things we did early on was clarify how mTBI needed to be defined. In this case because of our population of interest we were specifically using the definition provided by the VA and DoD and their clinical process guidelines. So that is one thing that limited some of the studies that met our inclusion criteria. A lot of the studies that we saw talked about mild TBI in very different ways with a lot of different definitions and we do include a table of studies in our final report that referred to Veterans and members of the military with mild TBI but did not use the VA DoD criteria for mild TBI. There are about 60 of those studies that were excluded from this report. So you can take a look at those in the final report if you are interested in those.
Anything that was not differentiating between like military and civilian populations or child and adult populations was excluded. Outcomes had to be things like health problems, cognitive deficits, functional limitations, mental health symptoms and cost and resource utilization. We looked at a variety of different study designs but as I said those ended up being pretty limited, basically due to what is feasible and ethical in this population.
Another criteria that we implemented was that there had to be a minimum of 30 mild TBI cases reported in each of those studies. Primarily that was because we wanted to be able to generalize as much as possible based on these results so when you are looking in observational studies that just has two, three, five, participants that have mild TBI, you cannot really say a lot in terms of generalizability. We chose 30 as our minimum sample size but this did exclude a few sort of promising current research topics that are being expanded in the future. One of the things to highlight is some of the CTE studies and other imaging related studies are generally smaller sample size studies and they got excluded from this report just because the samples were too small.
So just overview of results and then I am going to talk a bit about what strength of evidence means for these outcomes. Like I said we had 31 studies that met our inclusion criteria. Again for those of you who are not familiar with systematic review methodology, one of the main points of an evidence report or systematic review is to rate the strength of evidence overall. So to say here is the evidence that meets our scoping and inclusion criteria and then we want to talk about how strong is that evidence. Basically meaning how much do we have really solid faith in those results that if other studies were to come along in the future would they pretty likely to change our opinions or not so much.
In this case, because these studies are on a really complicated topic and it is really hard if not impossible to design more rigorous study designs like randomized control trials, then we rated the whole body of evidence as very low strength of evidence for all the outcomes being investigated in this report. Sometimes that was because there were only single studies that investigated certain outcomes. So when I was talking before about our key question two, which was looking at potential mediators or moderators of mTBI outcomes, so anything that might be a risk for protective factor, by and large there were very few studies that looked at any particular outcome. We could not even see a clear pattern for a lot of those results. Though there were some well conducted observational studies it is hard to put a ton of faith into any one single or few observational studies. So that is the main reason why the body of evidence is rated as very low strength.
So basically, as we are talking about the results here, and our stakeholders as well will be discussing this at the end, I want to emphasize and I am going to emphasize this over and over again that our findings are very tentative. So we are going to be talking about what we found. There were 31 studies so we do have a bunch of findings to talk about. However, because of the observational nature of these studies the findings are very, very tentative and we are going to emphasize that over and over again. Please do not go back to your clinical practices and think, okay well I know the hard and fast answer and there are no differences between these two populations or something like that because it is just not the case.
Let us see, in terms of overview of our results our cognitive, physical, mental health symptoms, we found that they were commonly reported by those with mTBI. However, one our very tentative findings, but findings that we found repeatedly, were that particularly for cognitive and mental health symptoms, they are not significantly more common in Veterans or members of the military with mTBI when you compare to a control group of very similar Veterans who have also been deployed but who do not have mild TBI. We are going to talk more about what that means in a specific results area. But again, very tentative results. This does not mean that mTBI does not exist. We are not saying anything like that. We are just saying that what the results of these studies show is that perhaps there are some Veteran and deployment specific factors that are really affecting folks when they come back from these conflicts recently. So we will talk about that in a bit. No consistent patterns of potential risk or protective factors were identified but we will highlight some of those results as we go.
To get into the specifics of the results the first section we found 17 studies on cognitive functioning that met our inclusion criteria. So an important thing to note about this, this is somewhat specific to cognitive functioning results and somewhat generalizable to all our results. Studies by and large reported mean scores. Like I said I am also a primary researcher. This is something that I have definitely done in the past and did not really think a lot about it until we were doing this report. When our stakeholders were saying, well we would like to know prevalence estimates. What are approximate percentages or proportions of Veterans with mTBI who are experiencing some of these difficulties? And when you have mean scores you cannot really tell that information. Particularly for cognitive functioning, mean scores are even more meaningless than in a lot of other areas. I say that because when you are defining cognitive functioning you are largely looking at change from baseline which is really hard to get a lot of times because it requires a longitudinal study design and so most of the studies were not able to look at change from baseline. In addition, mean scores should be reported… if you are reporting mean scores they should really be scores that are standardized scores. So at least some sort of estimates of impairment should be factored in if at all possible and by and large the studies did not do that. That is one thing for future research that would really be helpful to answer these kinds of questions. So for those of you doing research on mTBI looking at cognitive outcomes, that would be very helpful; reporting proportions of impaired scores and things like that.
That said, mean scores were within normal limits for all of the reported domains. So that includes language abilities, general funds of verbal knowledge, visual spatial abilities, memory, attention/concentration and executive functioning. However, there is a caveat about that and that is that those are mean scores. So that means that though many of the folks who are tested might have been performing within normal limits there can always be exceptions. So for those of you doing neuro-psych testing then please keep in mind that there can always be some exceptions and these are never hard and fast rules based on these results here.
Another thing to highlight from these results is that nonsignificant differences in cognitive performance compared to similar populations without mild traumatic brain injury was found in all reported domains again. So this is what I was talking about before that when you have a group of Veterans of members of the military who have mTBI and their scores on cognitive tests are compared to those very similar Veterans or members of the military who also experienced similar things like deployment. There were not statistically significant differences between those groups of Veterans with and without mTBI. These are on objective cognitive tests: a couple of things to point out is that there was significantly worse performance on some tests of memory, attention/concentration, and processing speed. But that was limited to… there were very few studies that believed… there were three studies authored by Cooper and Kennedy and Coldran. All three of those studies were using the ANAM and they were assessing active duty military members within 72 hours and then five days and then ten days following initial injury. Those statistically significant differences compared to a similar control group without mTBI disappeared after the first 72 hours to five days following injury.